1 / 27

Chapter 4

Chapter 4. The Von Neumann Model Basic components Instruction processing Memory. Building a computer. To get a task done by a computer, we need two things: a computer program that specifies what must be done to complete the task

yuki
Télécharger la présentation

Chapter 4

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Chapter 4 The Von Neumann Model Basic components Instruction processing Memory

  2. Building a computer • To get a task done by a computer, we need two things: • a computer program that specifies what must be done to complete the task • hardware that is capable of completing those tasks as specified • Low-level design of computer components is “digital design” • CEG 360 will cover much of this in detail. • Chapters 3 and 4 of the text will give you enough background to move forward to the next level of abstraction. • The basic computer model was proposed by John von Neumann in 1946

  3. Memory MAR MDR Processing Unit Input Output ALU TEMP (keyboard) (monitor) Control Unit PC IR The von Neumann Model - Illustration • Memory: holds both data and instructions • Processing Unit: carries out the instructions • Control Unit: sequences and interprets instructions • Input: external information into the memory • Output: produces results for the user

  4. Memory MAR MDR Processing Unit Input Output ALU TEMP (keyboard) (monitor) Control Unit PC IR Central Processing Unit (CPU)

  5. The von Neumann Model - CPU • Processing Unit - Does the actual work! • Can consist of many units, each specializing in one complex function. • At a minimum, has Arithmetic & Logic Unit (ALU) and temporary storage (registers). • The number of bits a basic Processing Unit operation can handle is called the WORD SIZE of the machine. • ALU • Performs basic operations: add, subtract, and, not, etc. • Generally operates on whole words of data. • Some can also operate on subsets of words (eg. single bits or bytes). • Registers: • Fast “on-board” storage for a small number of words. • Invaluable for intermediate data storage while processing. • Close to the ALU (much faster access than RAM). • General Purpose Registers (GPRs) are available to the programmer.

  6. 1Q 1D Q D Q /1Q CLR CLK 2Q 2D Q CLR D Q /2Q CLR 1D 1Q 2D 2Q 3Q 3D Q 3D 3Q D Q /3Q 4D 4Q CLR 4Q 4D Q D Q /4Q CLR CLK /CLR Registers: One way to hold binary information If R0 is 0011 Then R0[3:0] is 0011 R0[2:0] is 011 R0[2:1] is 01 R0[0:0] is 1

  7. CPU Overview Register File DR R0 3 R1 LdR R2 … SR2 SR1 R7 3 3 16 16 16

  8. Notations • Sets of Bits • A[3:0] denotes a set of 4 bits: A3, A2, A1, A0 • The content of an n-bit register R0 is referred to as R0[n-1:0] • R0n-1 is the most significant bit (MSB), or leftmost bit • R00 is the least significant bit (LSB), or rightmost bit • Given R0[31:0], R0[7:4] refers to the four bits from R4 to R7 • Bit Assignment • R3[5:0] ← R2[13:8] • Means that bits 5 to 0 of register R3 get assigned the values of bits 13 to 8 of register R2. • Contents • [R1] means “content of R1” • Mem[loc] means “content of memory location loc”

  9. Memory MAR MDR Processing Unit Input Output ALU TEMP (keyboard) (monitor) Control Unit PC IR Memory

  10. RAM Memory: A difference in scale • Memory is a large number of addressable fixed-sized locations • Each location has an address and contents • Address: bit pattern that uniquely identifies a memory location • Contents/Data: bit pattern stored at a given address. • analogy: p.o. boxes have fixed numbers, but changing contents. • A large number of locations • Each uniquely identifiable (its address) • Each having the ability to store a value of a certain size (its addressability) • Address Space n bits allow the addressing of 2n memory locations. • Example: 24 bits can address 224 = 16,777,216 locations (i.e. 16M locations)… but what SIZE is each location? ADD_IN DATA_IN DATA_OUT

  11. Memory Addressability • Word: the basic unit of data used by a device (generally the processing unit) • Ideally, one word is fetched with a single memory access • The addressability of Memory is the number of bits of memory referenced by a given address (generally in bytes) • Processor Word size for most contemporary computers is 32 bits • The memory may be byte OR word addressable. • Normally, a whole word is written and read at a time: • If the computer is word addressable, this is simply a single address location. • If the computer is byte addressable, and uses a multi-byte word, then the word address is conventionally either that of its most significant byte (big endian machines) or of its least significant byte (little endian machines).

  12. The von Neumann Model - Memory • We will access memory via a standard interface • The Memory Address Register sets up the decoder circuitry in the memory • The Memory Data Register hold contents to/from memory • The Write enable signal indicates which direction information should flow. • Memory Read • Read: the contents of the specified address will be written to the Memory Data Register. • If: Write enable is not applied in any given clock tick • Then: MDR ← M[(MAR)] • Memory Write • Write: the value to be stored is first written to the Memory Data Register, then the Write Enable is asserted, and the contents of the MDR are written to the specified address. • If: a Write enable is applied is any given clock tick • Then: M[(MAR)] ← MDR

  13. What is stored in memory? Is this value… • A positive integer? • A negative integer? • A float? • An address? • Data? • Part of a program (a machine instruction)? • How can we tell?

  14. The LC-3 so far… MAR 16 Register File DR 3 MDR 16 16 LdR 16 Memory Mem.EN R.W SR2 SR1 3 3 16 16

  15. Example problem • What sequence of control signals would need to be executed to add the value in R3 to the value in memory at address x30FF and place the result in R5? • You may assume that register R6 already contains the value x30FF

  16. Machine Words Sizes of C Objects (in Bytes) C Data Type Typical 32-bit • Int 4 • long int 4 • char 1 • short 2 • float 4 • double 8 • long double 8 • char * 4 • Or any other pointer • Machine Has “Word Size” • Nominal size of integer-valued data • Including addresses • Most current machines are 32 bits (4 bytes) • Limits addresses to 4GB • Becoming too small for memory-intensive applications • High-end systems are 64 bits (8 bytes) • Potentially address  1.8 X 1019 bytes • Machines support multiple data formats • Fractions or multiples of word size • Always integral number of bytes

  17. 0x100 0x101 0x102 0x103 01 23 45 67 0x100 0x101 0x102 0x103 67 45 23 01 Byte Ordering Example • How should bytes within multi-byte word be ordered in memory? • Big Endian • Least significant byte has highest address • Little Endian • Least significant byte has lowest address • Example • Variable x has 4-byte representation 0x01234567 • Address given by &x is 0x100 Big Endian 01 23 45 67 Little Endian 67 45 23 01

  18. Memory MAR MDR Processing Unit Input Output ALU TEMP (keyboard) (monitor) Control Unit PC IR Memory

  19. The von Neumann Model – Control Unit and I/O • Control Unit • The control unit coordinates all actions needed to execute the instruction • It fetches & decodes each instruction, and sets up the appropriate inputs for the Memory, Processing, and I/O units as required. • It communicates with memory via the Program Counter (PC) and Instruction Register (IR) • PC (aka Instruction Pointer) • Holds the address of the next instruction to be fetched. • IR • Holds the instruction currently being executed. • This can be a single word, or multiple words. • Input & Output • Generically known as peripherals - they are external to the CPU. • This means we will have to develop mechanisms for autonomous devices to communicate with each other - more on this later.

  20. The LC-3: A von Neumann Machine

  21. Instruction Cycle - overview • The Control Unit orchestrates the complete execution of each instruction: • At its heart is a Finite State Machine that sets up the state of the logic circuits according to each instruction. • This process is governed by the system clock - the FSM goes through one transition (“machine cycle”) for each tick of the clock. • Six phases of the complete Instruction Cycle (and repeat…) • Fetch: load IR with instruction from memory • Decode: determine action to take (set up inputs for ALU, RAM, etc.) • Evaluate address: compute memory address of operands, if any • Fetch operands: read operands from memory or registers • Execute: carry out instruction • Store results: write result to destination (register or memory)

  22. Clock cycles • One clock cycle is often (though not always) the time it takes for the machine to perform a simple instruction (such as a register-to-register add) • Usually one clock cycle corresponds to a complete instruction cycle • The clock cycle is broken down into sub-cycles for each phase • The control unit executes the appropriate signals for each phase: if (instruction is add && sub-cycle is 4) { SR1 source register number etc… }

  23. Instruction Cycle - steps 1 & 2 • Fetch • The first step is to read an instruction from memory. • This actually takes several smaller steps, or “micro-instructions”: • MAR  (PC) ; use the value in PC to access memory • PC  (PC) + 1 ; increment the value of PC • MDR  Mem[MAR] ; read memory location to MDR • IR  (MDR) ; copy (MDR) to IRDecode • Steps 1, 2 & 4 take a single machine cycle each, but step 3 (memory access) can take many machine cycles. • Decode • The opcode is input to a decoder, which sets up the ensuing sequence of events according the instruction.

  24. Instruction Cycle - steps 3 & 4 • Evaluate Address • Computes the address of the operand (if any), or of the memory location to be accessed: e.g. the location from which to obtain a value in a LOAD instruction. • This is known as the Effective Address (EA). • Register to Register instructions don’t have an evaluate address phase. • Fetch Operands • Obtains the source operand(s), if required for execution. • Operands can come from Registers or RAM, or be embedded in the instruction itself. • The Effective Address (EA) determined in the previous step may be used to obtain an operand from memory.

  25. Instruction Cycle - steps 5 & 6 • Execute • Now that everything is in place, the instruction is executed. • e.g. if the opcode was ADD, the two source operands are added by the ALU. • If the opcode was a control instruction, a value is written to the PC • Data Movement instructions don’t have an execute phase • Store Result • If there is a result from the operation it is written to memory (using the EA), or to a register. • Then REPEAT • The control unit just keeps repeating this whole process: Recall that the PC was incremented in the first step (FETCH), so the instruction retrieved will be the next in the program as stored in memory - unless the instruction just executed changed the contents of the PC.

  26. Types of Instruction • Operate Instructions • process data (addition, logical operations, etc.) • Data Movement Instructions … • move data between memory locations and registers. • Control Instructions … • change the sequence of execution of instructions in the stored program. • The default is sequential execution: the PC is incremented by 1 at the start of every Fetch, in preparation for the next one. • Control instructions set the PC to a new value during the Execute phase, so the next instruction comes from a different place in the program. • This allows us to build control structures such as loops and branches.

  27. Problems • 4.5, 4.7, 4.8, 4.10

More Related